Update config.json

config.json

@@ -1,51 +1,50 @@
-{
-  "_name_or_path": "answerdotai/ModernBERT-base",
-  "architectures": [
-    "ModernBergForMaskedLM"
-  ],
-  "attention_bias": false,
-  "attention_dropout": 0.0,
-  "bos_token_id": 50281,
-  "classifier_activation": "gelu",
-  "classifier_bias": false,
-  "classifier_dropout": 0.0,
-  "classifier_pooling": "mean",
-  "cls_token_id": 50281,
-  "conv1d_width": 4,
-  "decoder_bias": true,
-  "deterministic_flash_attn": false,
-  "embedding_dropout": 0.0,
-  "eos_token_id": 50282,
-  "global_attn_every_n_layers": 3,
-  "global_rope_theta": 160000.0,
-  "global_temporal_every_n_layers": 3,
-  "gradient_checkpointing": false,
-  "hidden_activation": "gelu",
-  "hidden_size": 768,
-  "initializer_cutoff_factor": 2.0,
-  "initializer_range": 0.02,
-  "intermediate_size": 1152,
-  "layer_norm_eps": 1e-05,
-  "local_attention": 128,
-  "local_rope_theta": 10000.0,
-  "lru_width": 1152,
-  "max_position_embeddings": 8192,
-  "mlp_bias": false,
-  "mlp_dropout": 0.0,
-  "model_type": "modernbert",
-  "norm_bias": false,
-  "norm_eps": 1e-05,
-  "num_attention_heads": 12,
-  "num_hidden_layers": 22,
-  "pad_token_id": 50283,
-  "position_embedding_type": "absolute",
-  "reference_compile": null,
-  "repad_logits_with_grad": false,
-  "rope_theta": 10000.0,
-  "sep_token_id": 50282,
-  "sparse_pred_ignore_index": -100,
-  "sparse_prediction": false,
-  "torch_dtype": "float32",
-  "transformers_version": "4.48.0",
-  "vocab_size": 50368
-}
+{
+  "_name_or_path": "answerdotai/ModernBERT-base",
+  "architectures": [
+    "ModernBergForMaskedLM"
+  ],
+  "auto_map": {
+    "AutoConfig": "modernberg_model.ModernBergConfig",
+    "AutoModel": "modernberg_model.ModernBergModel",
+    "AutoModelForMaskedLM": "modernberg_model.ModernBergModelForMaskedLM",
+  },
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "global_rope_theta": 160000.0,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "layer_norm_eps": 1e-05,
+  "local_attention": 128,
+  "local_rope_theta": 10000.0,
+  "lru_width": 1152,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 50283,
+  "position_embedding_type": "absolute",
+  "sep_token_id": 50282,
+  "tie_word_embeddings": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.47.0.dev0",
+  "vocab_size": 50368
+}